EP1378853A1 - Digital medical assistance system - Google Patents

Digital medical assistance system Download PDF

Info

Publication number
EP1378853A1
EP1378853A1 EP02077665A EP02077665A EP1378853A1 EP 1378853 A1 EP1378853 A1 EP 1378853A1 EP 02077665 A EP02077665 A EP 02077665A EP 02077665 A EP02077665 A EP 02077665A EP 1378853 A1 EP1378853 A1 EP 1378853A1
Authority
EP
European Patent Office
Prior art keywords
data
medical
patient
rules
memory
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Withdrawn
Application number
EP02077665A
Other languages
German (de)
French (fr)
Inventor
Serge Muller
Guillaume Crepin
Elisabeth Soubelet
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
GE Medical Systems Global Technology Co LLC
Original Assignee
GE Medical Systems Global Technology Co LLC
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by GE Medical Systems Global Technology Co LLC filed Critical GE Medical Systems Global Technology Co LLC
Priority to EP02077665A priority Critical patent/EP1378853A1/en
Priority to JP2003270543A priority patent/JP4651271B2/en
Priority to JP2003192106A priority patent/JP2004049909A/en
Publication of EP1378853A1 publication Critical patent/EP1378853A1/en
Withdrawn legal-status Critical Current

Links

Images

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
    • G16H10/60—ICT specially adapted for the handling or processing of patient-related medical or healthcare data for patient-specific data, e.g. for electronic patient records
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H15/00—ICT specially adapted for medical reports, e.g. generation or transmission thereof
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H30/00—ICT specially adapted for the handling or processing of medical images
    • G16H30/20—ICT specially adapted for the handling or processing of medical images for handling medical images, e.g. DICOM, HL7 or PACS
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H40/00—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices
    • G16H40/60—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices for the operation of medical equipment or devices
    • G16H40/63—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices for the operation of medical equipment or devices for local operation

Definitions

  • the invention relates to a digital medical assistance system.
  • Data can also be observations with descriptive parameters from analyzes of the patient's medical imagery, the images being acquired according to different processes (e.g. the form of microcalcifications in X-ray mammograms, the over-densities appearing on images taken from magnetic resonance with the use of contrast media, shape and texture of ultrasound ultrasound signals).
  • Data can also come from a clinical examination carried out on the patient, from recordings metabolic (e.g. electrocardiogram, encephalogram) or analysis cytological or histological of cells or tissues taken from the patient.
  • Document EP-A-1035507 describes a detection assistance system computer-aided (in English Computer-Aided Detection or CAD). This system allows you to read a medical image and analyze it to extract suspicious areas reflecting the presence of a lesion. Only information quantitative data on lesions are provided.
  • CAD Computer-Aided Detection
  • a digital system medical assistance for decision-making includes a reception interface receiving clinical data and physical registration data or physiological.
  • the system also includes a rule storage memory medical and a detection assistance unit receiving the registration data, the unit providing data highlighting a pathology.
  • a motor inference in the system receives clinical data from the interface, rules medical rules storage memory, and the data provided by the detection assistance unit. Engine provides recommendation to take of decision.
  • the system has the following advantages.
  • the system provides a recommendation to the practitioner on the decision he can take to continue the patient study process.
  • the system takes into account all the data concerning a patient and the rules used by practitioners to take a decision.
  • the recommendation provided by the system prevents, for example, that the practitioner has recourse to an operation of the patient, at too early a stage in the patient study process, and that he prefers to first take the organ tissue samples by a less invasive means.
  • Figure 1 shows a schematic view of an assistance system Medical.
  • FIG. 1 shows a schematic view of the digital system.
  • the system 10 presents a memory 18 for storing medical rules and a reception interface 12 receiving clinical data 14 and physical or physiological recording data 16 concerning the patient.
  • the system 10 also includes a detection aid unit 20; Single 20 receives registration data 16 and outputs data 17 highlighting a pathology.
  • An inference engine 22 in the system 10 receives clinical data 14, medical rules of memory 18 as well as the data 17 supplied by the detection aid unit 20. Motor 22 provides a recommendation for decision making.
  • the reception interface 12 receives clinical data 14 on the patient who are objective information about the patient. They include information on the patient's height, weight or age. Clinical data 14 also include information about the patient's clinical history, such as than diseases previously contracted by the patient. Data 14 are for example collected using a network interface with a questionnaire which can be filled in by the patient; data 14 can also be collected by the practitioner who connects to a network so that he can receive data from other computers.
  • the interface 12 also receives the physical recording data 16 or physiological of the patient.
  • Data 16 is also information about patient but which are subject to varying interpretations depending on the practitioner.
  • the data 16 can be for example observations made by the practitioner on the patient during physical exams.
  • the data can also be data from a medical imaging device. Imaging device delivers for example images with 2D, 3D format (volume, image and time) or 4D (e.g. volume and time). Images can be provided by several imaging devices, which deliver images using different properties of the interaction between radiation and lesions.
  • the images are for example images from CT or radiological or ultrasound images that provide a morphological description of the patient; images can also come from nuclear medicine devices, functional MRI or positron emission tomography.
  • the data is stored in a database, for example in RAM memory. At each stage of the patient's study, the practitioner can enter new data 14, 16.
  • the storage memory 18 stores the practitioner's medical rules.
  • the medical rules are used by practitioners or medical experts in their studies of disease families. Medical rules are a model of decision that the practitioner applies to make his decisions.
  • Memory 18 of rule storage can include the user practitioner's decision model, established on the basis of its own expertise accumulated by its experience. Of advantageously, the rules can also be rules established by an expert of the medical field concerned.
  • the system 10 has the advantage of allowing an inexperienced practitioner to use the system 10 recommendation as a guide and take advantage of expert knowledge.
  • Rules established by a plurality of medical experts can also be integrated into memory 18. The practitioner can then obtain a recommendation by selecting a type of medical rules. The practitioner can also obtain a plurality of recommendations that he can compare with each other.
  • the detection assistance unit 20 receives input 16; at output, unit 20 provides data 17 highlighting the existence of a pathology.
  • Data 16 is information about the patient and on which the practitioner relies for make a decision to take the next step in the review process patient.
  • the data 17 provided at the output are quantitative measurements on the existence of pathologies. For example, when the data 16 input come from data provided by a medical imaging device, unit 20 reads images, analyzes and extracts suspicious areas reflecting the presence of a lesion. Data 17 may include a malignancy index reflecting the risk that the lesion is cancerous. Data 17 can also understand a location of lesions on a patient's organ. Unit 20 thus makes it possible to synthesize the data provided for example by imagery and assist the practitioner to improve the quality of detection or characterization during image analysis. Unit 20 allows system 10 to take clinical data 14 into consideration together with data 16 recording.
  • the inference engine 22 provides a recommendation to the practitioner for assist him in his decision-making.
  • the inference engine exploits the base information including data describing the patient. Those data include clinical data 14 received from interface 12 and data 17 supplied by unit 20.
  • the motor 22 also receives the medical rules of the memory 18; from medical rules, and based on data 14 and 17 the Motor 22 simulates the reasoning and decisions of the practitioner.
  • Engine of inference 22 is for example of order zero: it provides a recommendation to the practitioner by performing the practitioner's reasoning.
  • the inference engine 22 can also be of order 1; motor 22 reconsiders the whole process tree decision-making to validate the consideration of a hypothesis that was not not most likely at an earlier stage in the patient's examination.
  • the practitioner may be required to consider several hypotheses of pathologies with degrees of malignancy which lead to as many therapeutic decisions.
  • the course of data up to clinical hypotheses can be represented as a tree of decision. Additional data can modify the number of hypotheses to consider and the relative weight of these assumptions.
  • the decision tree used previously is then modified.
  • Other technologies such as fuzzy trees can be chosen for the implementation of the rules.
  • Neural networks can also be used to build a decision-making process tree based on knowledge of data collected during the exam.
  • the recommendation provided by the engine 22 of the system 10 is a indication for the practitioner of the next step which can be applied to the patient.
  • the recommendation provided by system 10 optimizes the probability that the practitioner's decision has a positive impact on the patient's health.
  • the recommendation is for example the invitation to take new images medical information for further patient data, needle biopsy, resort to a surgical stage, or the administration of specific drugs (chemotherapy).
  • System 10 can supplement the recommendation with an index confidence or probability of effectiveness.
  • the index is for example established by the inference engine 22 during the different stages of reasoning simulation by a tree process.
  • the system can also provide a plurality of recommendations by issuing for each recommendation a probability efficiency.
  • the recommendation provided by the system reduces the variations between decisions made by different practitioners in the study of the patient.
  • the digital system 10 is for example a computer.
  • the system 10 further comprises an interface 24 enrichment in medical rules of memory 18 for storing rules medical.
  • Medical rules can be stored on memory 18 before to use system 10 on several patients; memory 18 can also be updated with new medical rules at each stage of the study of a patient.
  • the new rules are for example loaded from a network or at from archiving bodies such as a CD-ROM. This can be achieved, by example, using a network interface that allows introduction of assumptions, rules and conclusions in natural language or which instructs them to from recording media (eg floppy disks or CD-ROMs).
  • An interpreter provides input translation into natural language into rules which can be executed by a computer (for example using a language such than Prolog).
  • the system may further include a rule deduction module 26 medical from the comparison between the truth about the pathology obtained has posteriori - after anatomo-pathological analysis for example - and the data in input of interface 12 and output of unit 20, module 26 being able to supply medical rules in memory 18 for storing medical rules.
  • a rule deduction module 26 medical from the comparison between the truth about the pathology obtained has posteriori - after anatomo-pathological analysis for example - and the data in input of interface 12 and output of unit 20, module 26 being able to supply medical rules in memory 18 for storing medical rules.
  • a supervised learning mechanism which, from patient data and the truth obtained a posteriori, allows to update the decision rules of the tree decision structure.
  • we can use an artificial perceptron neural network for which learning makes it possible to update the weight of its connections.
  • the news constructed rules supplement the memory 18 for storing medical rules.
  • the system finds, for example, an application in mammography.

Abstract

The digital assistant has an interface (12) to receive clinical (14) and physiological (16) data, a memory (18) holding medical rules and an inference engine (22) to process data using the rules. The unit accepts the input data and then delivers a diagnostic recommendation. An interface (24) is provided to allow enrichment of the rule set.

Description

L'invention concerne un système numérique d'assistance médicale.The invention relates to a digital medical assistance system.

Les connaissances en radiologie, et en médecine en général, sont difficiles à acquérir et sont souvent partielles et floues dans les domaines les plus difficiles. Il a été observé que des conclusions différentes ont pu être formulées par des praticiens au sujet de patients présentant des symptômes et des antécédents similaires. Ces différences proviennent de l'importante quantité de données sur les patients que les praticiens ont à considérer ; ces différences proviennent aussi des incertitudes liées aux critères de décision des praticiens. Pour une spécialité médicale donnée, seuls quelques experts sont capables, à partir de recherches poussées, de déterminer et d'utiliser avec de fortes chances de succès des modèles de décision. Ces modèles de décision indiquent quelle procédure appliquer au patient à partir d'une série de données collectée et analysée. Ces données sont par exemple des réponses à des questionnaires (par ex. un historique familial ou personnel, des habitudes ou des symptômes du patient). Les données peuvent aussi être des observations comportant des paramètres descriptifs provenant d'analyses d'imageries médicales du patient, les images étant acquises selon différents procédés (par ex. la forme de microcalcifications dans des mammographies par rayons X, les sur-densités apparaissant sur des images issues de résonance magnétique avec utilisation de produits de contraste , la forme et la texture de signaux échographiques en ultrasons). Les données peuvent aussi provenir d'un examen clinique effectué sur le patient, d'enregistrements métaboliques (ex : électrocardiogramme, encéphalogramme) ou d'analyse cytologique ou histologique de cellules ou de tissus prélevés sur le patient.Knowledge of radiology, and medicine in general, is difficult to acquire and are often partial and unclear in the most difficult areas. he has been observed that different conclusions may have been reached by practitioners about patients with symptoms and history Similar. These differences are due to the large amount of data on patients that practitioners have to consider; these differences also come from uncertainties related to practitioners' decision criteria. For a specialty given medical data, only a few experts are capable, based on research pushes, determine and use models with high chances of success of decision. These decision models indicate which procedure to apply to patient from a series of data collected and analyzed. These data are by example of responses to questionnaires (e.g. family history or personal, habits or symptoms of the patient). Data can also be observations with descriptive parameters from analyzes of the patient's medical imagery, the images being acquired according to different processes (e.g. the form of microcalcifications in X-ray mammograms, the over-densities appearing on images taken from magnetic resonance with the use of contrast media, shape and texture of ultrasound ultrasound signals). Data can also come from a clinical examination carried out on the patient, from recordings metabolic (e.g. electrocardiogram, encephalogram) or analysis cytological or histological of cells or tissues taken from the patient.

Au cours de l'examen de patients, la plupart des praticiens sont en mesure d'obtenir des performances élevées dans leur processus décisionnel, lorsque le cas du patient est typique à diagnostiquer ou lorsque la procédure est unique et largement admise. En revanche, lorsque le cas du patient est atypique ou lorsque le médecin est inexpérimenté, il est parfois difficile de prendre une décision à une étape donnée dans le processus d'examen du patient. Il est souvent fait appel à des experts dans des cas délicats, pour confirmer ou suggérer une décision et l'étape suivante à suivre dans le processus d'examen du patient. Le nombre limité d'experts réduit les possibilités du recours à leur service. De plus, l'avis des experts eux-mêmes est susceptible de varier.During the examination of patients, most practitioners are able obtain high performance in their decision-making process, when the case of the patient is typical to diagnose or when the procedure is unique and widely accepted. However, when the patient's case is atypical or when the doctor is inexperienced, it is sometimes difficult to make a decision at a given step in the patient examination process. Frequently, experts in delicate cases, to confirm or suggest a decision and step next to follow in the patient review process. The limited number of experts reduces the possibilities of using their service. In addition, the opinion of experts themselves is likely to vary.

Le document EP-A-1035507 décrit un système d'assistance à la détection assistée par ordinateur (en anglais Computer-Aided Détection ou CAD). Ce système permet de lire une image médicale et de l'analyser pour en extraire des zones suspicieuses traduisant la présence d'une lésion. Seules des informations quantitatives sur les lésions sont fournies.Document EP-A-1035507 describes a detection assistance system computer-aided (in English Computer-Aided Detection or CAD). This system allows you to read a medical image and analyze it to extract suspicious areas reflecting the presence of a lesion. Only information quantitative data on lesions are provided.

Il existe donc un besoin d'assister les praticiens à prendre des décisions lors de l'étude d'un patient alors qu'ils ont à considérer une multitude des données concernant le patient.There is therefore a need to assist practitioners in making decisions when of the study of a patient while they have to consider a multitude of data concerning the patient.

BREF RESUME DE L'INVENTIONBRIEF SUMMARY OF THE INVENTION

Selon un mode de réalisation l'invention, un système numérique d'assistance médicale à la prise de décision comprend une interface de réception recevant des données cliniques et des données d'enregistrement physiques ou physiologiques. Le système comprend aussi une mémoire de stockage de règles médicales et une unité d'aide à la détection recevant les données d'enregistrement, l'unité fournissant des données mettant en évidence une pathologie. Un moteur d'inférences dans le système reçoit les données cliniques de l'interface, les règles médicales de la mémoire de stockage des règles médicales, et les données fournies par l'unité d'aide à la détection. Le moteur fournit une recommandation à la prise de décision.According to one embodiment of the invention, a digital system medical assistance for decision-making includes a reception interface receiving clinical data and physical registration data or physiological. The system also includes a rule storage memory medical and a detection assistance unit receiving the registration data, the unit providing data highlighting a pathology. A motor inference in the system receives clinical data from the interface, rules medical rules storage memory, and the data provided by the detection assistance unit. Engine provides recommendation to take of decision.

Le système présente les avantages suivants. Le système fournit une recommandation au praticien sur la décision qu'il peut prendre pour poursuivre le processus d'étude du patient. Le système prend en compte toutes les données concernant un patient et les règles utilisées par les praticiens pour prendre une décision. La recommandation fournie par le système évite, par exemple, que le praticien ait recours à une opération du patient, à un stade trop précoce dans le processus d'étude du patient, et qu'il préfère prélever dans un premier temps les échantillons de tissu d'organes par un moyen moins invasif.The system has the following advantages. The system provides a recommendation to the practitioner on the decision he can take to continue the patient study process. The system takes into account all the data concerning a patient and the rules used by practitioners to take a decision. The recommendation provided by the system prevents, for example, that the practitioner has recourse to an operation of the patient, at too early a stage in the patient study process, and that he prefers to first take the organ tissue samples by a less invasive means.

BREVE DESCRIPTION DES DESSINSBRIEF DESCRIPTION OF THE DRAWINGS

La figure 1 montre une vue schématique d'un système d'assistance médicale.Figure 1 shows a schematic view of an assistance system Medical.

DESCRIPTION DETAILLEE DE L'INVENTIONDETAILED DESCRIPTION OF THE INVENTION

En référence à la figure 1, est décrit un système 10 numérique d'assistance médicale à la prise de décision. La figure 1 montre une vue schématique du système numérique. Le système 10 présente une mémoire 18 de stockage de règles médicales et une interface 12 de réception recevant des données cliniques 14 et des données d'enregistrement 16 physiques ou physiologiques concernant le patient. Le système 10 comprend aussi une unité 20 d'aide à la détection ; l'unité 20 reçoit les données d'enregistrement 16 et fournit en sortie des données 17 mettant en évidence une pathologie. Un moteur d'inférences 22 dans le système 10 reçoit les données cliniques 14, les règles médicales de la mémoire 18 ainsi que les données 17 fournies par l'unité 20 d'aide à la détection. Le moteur 22 fournit une recommandation à la prise de décision.With reference to FIG. 1, a digital assistance system 10 is described. medical decision-making. Figure 1 shows a schematic view of the digital system. The system 10 presents a memory 18 for storing medical rules and a reception interface 12 receiving clinical data 14 and physical or physiological recording data 16 concerning the patient. The system 10 also includes a detection aid unit 20; Single 20 receives registration data 16 and outputs data 17 highlighting a pathology. An inference engine 22 in the system 10 receives clinical data 14, medical rules of memory 18 as well as the data 17 supplied by the detection aid unit 20. Motor 22 provides a recommendation for decision making.

L'interface 12 de réception reçoit les données cliniques 14 sur le patient qui sont des informations objectives sur le patient. Elles comprennent des informations sur la taille, le poids ou l'âge du patient. Les données cliniques 14 comprennent aussi des informations sur les antécédents cliniques du patient, telles que des maladies antérieurement contractées par le patient. Les données 14 sont par exemple collectées en utilisant une interface de réseau avec un questionnaire qui peut être rempli par le patient ; les données 14 peuvent aussi être collectées par le praticien qui se connecte à un réseau de telle sorte qu'il lui est possible de recevoir des données d'autres ordinateurs.The reception interface 12 receives clinical data 14 on the patient who are objective information about the patient. They include information on the patient's height, weight or age. Clinical data 14 also include information about the patient's clinical history, such as than diseases previously contracted by the patient. Data 14 are for example collected using a network interface with a questionnaire which can be filled in by the patient; data 14 can also be collected by the practitioner who connects to a network so that he can receive data from other computers.

L'interface 12 reçoit par ailleurs les données d'enregistrement 16 physiques ou physiologique du patient. Les données 16 sont également des informations sur le patient mais qui sont sujettes à des interprétations variables selon les praticiens. Les données 16 peuvent être par exemple des observations faites par le praticien sur le patient durant des examens physiques. Les données peuvent aussi être des données issues d'un appareil d'imagerie médicale. L'appareil d'imagerie délivre par exemple des images avec un format 2D, 3D (volume, image et temps) ou 4D (par ex. volume et temps). Des images peuvent être fournies par plusieurs appareils d'imagerie, qui délivrent des images en utilisant différentes propriétés de l'interaction entre radiation et lésions. Les images sont par exemple des images issues de scanner ou d'images radiologiques ou échographiques qui fournissent une description morphologique du patient ; les images peuvent également provenir d'appareils de médecine nucléaire, d'IRM fonctionnelle ou de tomographie à émission de positrons.The interface 12 also receives the physical recording data 16 or physiological of the patient. Data 16 is also information about patient but which are subject to varying interpretations depending on the practitioner. The data 16 can be for example observations made by the practitioner on the patient during physical exams. The data can also be data from a medical imaging device. Imaging device delivers for example images with 2D, 3D format (volume, image and time) or 4D (e.g. volume and time). Images can be provided by several imaging devices, which deliver images using different properties of the interaction between radiation and lesions. The images are for example images from CT or radiological or ultrasound images that provide a morphological description of the patient; images can also come from nuclear medicine devices, functional MRI or positron emission tomography.

Les données sont stockées dans une banque de données, par exemple dans une mémoire vive RAM. A chaque étape de l'étude du patient, le praticien peut entrer de nouvelles données 14, 16.The data is stored in a database, for example in RAM memory. At each stage of the patient's study, the practitioner can enter new data 14, 16.

La mémoire 18 de stockage stocke les règles médicales du praticien. Les règles médicales sont utilisées par les praticiens ou les experts médicaux dans leurs études de familles de maladies. Les règles médicales sont un modèle de décision que le praticien applique pour prendre ses décisions. La mémoire 18 de stockage de règles peut comprendre le modèle de décision du praticien utilisateur, établi sur la base de sa propre expertise accumulée par son expérience. De manière avantageuse, les règles peuvent aussi être des règles établies par un expert du domaine médical concerné. Le système 10 présente l'avantage de permettre à un praticien inexpérimenté d'utiliser la recommandation du système 10 comme un guide et de profiter des connaissances d'experts.The storage memory 18 stores the practitioner's medical rules. The medical rules are used by practitioners or medical experts in their studies of disease families. Medical rules are a model of decision that the practitioner applies to make his decisions. Memory 18 of rule storage can include the user practitioner's decision model, established on the basis of its own expertise accumulated by its experience. Of advantageously, the rules can also be rules established by an expert of the medical field concerned. The system 10 has the advantage of allowing an inexperienced practitioner to use the system 10 recommendation as a guide and take advantage of expert knowledge.

Des règles instituées par une pluralité d'experts médicaux peuvent aussi être intégrées dans la mémoire 18. Le praticien peut alors obtenir une recommandation en sélectionnant un type de règles médicales. Le praticien peut aussi obtenir une pluralité de recommandations qu'il peut comparer entre elles.Rules established by a plurality of medical experts can also be integrated into memory 18. The practitioner can then obtain a recommendation by selecting a type of medical rules. The practitioner can also obtain a plurality of recommendations that he can compare with each other.

L'unité 20 d'aide à la détection (en anglais Computer-Aided Détection ou CAD) reçoit en entrée les données 16 ; en sortie, l'unité 20 fournit les données 17 mettant en évidence l'existence d'une pathologie. Les données 16 sont des informations concernant le patient et sur lesquelles le praticien s'appuie pour prendre une décision pour passer à une prochaine étape du processus d'examen du patient. Les données 17 fournies en sortie sont des mesures quantitatives sur l'existence de pathologies. Par exemple, lorsque les données 16 fournies en entrée sont issues de données fournies par un appareil d'imagerie médicale, l'unité 20 lit les images, les analyse et en extrait des zones suspicieuses traduisant la présence d'une lésion. Les données 17 peuvent comprendre un index de malignité traduisant le risque que la lésion soit cancéreuse. Les données 17 peuvent aussi comprendre une localisation de lésions sur l'organe d'un patient. L'unité 20 permet ainsi de synthétiser les données fournies par exemple par l'imagerie médicale et d'assister le praticien pour améliorer la qualité de la détection ou de la caractérisation lors de l'analyse des images. L'unité 20 permet au système 10 de prendre les données cliniques 14 en considération ensemble avec les données 16 d'enregistrement.The detection assistance unit 20 (in English Computer-Aided Detection or CAD) receives input 16; at output, unit 20 provides data 17 highlighting the existence of a pathology. Data 16 is information about the patient and on which the practitioner relies for make a decision to take the next step in the review process patient. The data 17 provided at the output are quantitative measurements on the existence of pathologies. For example, when the data 16 input come from data provided by a medical imaging device, unit 20 reads images, analyzes and extracts suspicious areas reflecting the presence of a lesion. Data 17 may include a malignancy index reflecting the risk that the lesion is cancerous. Data 17 can also understand a location of lesions on a patient's organ. Unit 20 thus makes it possible to synthesize the data provided for example by imagery and assist the practitioner to improve the quality of detection or characterization during image analysis. Unit 20 allows system 10 to take clinical data 14 into consideration together with data 16 recording.

Le moteur d'inférence 22 fournit une recommandation au praticien pour l'assister dans sa prise de décision. Le moteur d'inférence exploite la base d'informations comprenant les données décrivant le patient. Ces données comprennent les données cliniques 14 reçues de l'interface 12 et les données 17 fournies par l'unité 20. Le moteur 22 reçoit aussi les règles médicales de la mémoire 18 ; à partir des règles médicales, et sur la base des données 14 et 17 le moteur 22 simule les raisonnements et les décisions du praticien. Le moteur d'inférence 22 est par exemple d'ordre zéro : il fournit une recommandation au praticien en effectuant le raisonnement du praticien. Le moteur d'inférence 22 peut aussi être d'ordre 1 ; le moteur 22 reconsidère l'ensemble du processus décisionnel arborescent pour valider la considération d'une hypothèse qui n'était pas la plus probable lors d'une étape antérieure dans l'examen du patient. A partir d'un ensemble de données acquises sur le patient, le praticien peut être amené à considérer plusieurs hypothèses de pathologies avec des degrés de malignité différents qui conduisent à autant de décision thérapeutiques. Le parcours des données jusqu'aux hypothèses cliniques peut être représenté sous forme d'un arbre de décision. Une donnée supplémentaire peut modifier le nombre d'hypothèses à considérer et le poids relatif de ces hypothèses. L'arbre de décision utilisé précédemment est alors modifié. D'autres technologies telles que les arbres flous peuvent être choisies pour l'implémentation des règles. Des réseaux de neurones artificiels peuvent aussi être utilisés pour construire un processus décisionnel arborescent à partir de la connaissance de données collectées pendant l'examen.The inference engine 22 provides a recommendation to the practitioner for assist him in his decision-making. The inference engine exploits the base information including data describing the patient. Those data include clinical data 14 received from interface 12 and data 17 supplied by unit 20. The motor 22 also receives the medical rules of the memory 18; from medical rules, and based on data 14 and 17 the Motor 22 simulates the reasoning and decisions of the practitioner. Engine of inference 22 is for example of order zero: it provides a recommendation to the practitioner by performing the practitioner's reasoning. The inference engine 22 can also be of order 1; motor 22 reconsiders the whole process tree decision-making to validate the consideration of a hypothesis that was not not most likely at an earlier stage in the patient's examination. From of a set of data acquired on the patient, the practitioner may be required to consider several hypotheses of pathologies with degrees of malignancy which lead to as many therapeutic decisions. The course of data up to clinical hypotheses can be represented as a tree of decision. Additional data can modify the number of hypotheses to consider and the relative weight of these assumptions. The decision tree used previously is then modified. Other technologies such as fuzzy trees can be chosen for the implementation of the rules. Neural networks can also be used to build a decision-making process tree based on knowledge of data collected during the exam.

La recommandation fournie par le moteur 22 du système 10 est une indication pour le praticien de la prochaine étape qui peut être appliquée sur le patient. La recommandation fournie par le système 10 optimalise la probabilité que la décision du praticien ait un impact positif sur la santé du patient. La recommandation est par exemple l'invitation à effectuer de nouvelles images médicales pour obtenir d'autres données sur le patient, une biopsie à l'aiguille, à recourir à une étape de chirurgie, ou l'administration de drogues spécifiques (chimiothérapie). Le système 10 peut compléter la recommandation par un index de confiance ou de probabilité d'efficacité. L'index est par exemple établi par le moteur d'inférence 22 lors des différentes étapes de la simulation du raisonnement par un processus arborescent. Le système peut aussi fournir une pluralité de recommandations en délivrant pour chacune des recommandations une probabilité d'efficacité. La recommandation fournie par le système réduit les variations entre les décisions prises par différents praticiens dans l'étude du patient.The recommendation provided by the engine 22 of the system 10 is a indication for the practitioner of the next step which can be applied to the patient. The recommendation provided by system 10 optimizes the probability that the practitioner's decision has a positive impact on the patient's health. The recommendation is for example the invitation to take new images medical information for further patient data, needle biopsy, resort to a surgical stage, or the administration of specific drugs (chemotherapy). System 10 can supplement the recommendation with an index confidence or probability of effectiveness. The index is for example established by the inference engine 22 during the different stages of reasoning simulation by a tree process. The system can also provide a plurality of recommendations by issuing for each recommendation a probability efficiency. The recommendation provided by the system reduces the variations between decisions made by different practitioners in the study of the patient.

Le système 10 numérique est par exemple un ordinateur.The digital system 10 is for example a computer.

Avantageusement, le système 10 comprend en outre une interface 24 d'enrichissement en règles médicales de la mémoire 18 de stockage de règles médicales. Les règles médicales peuvent être stockées sur la mémoire 18 avant d'utiliser le système 10 sur plusieurs patients ; la mémoire 18 peut aussi être remise à jour avec de nouvelles règles médicales à chaque étape de l'étude d'un patient. Les nouvelles règles sont par exemple chargées depuis un réseau ou à partir d'organes d'archivage tels que un CD-ROM. Ceci peut être réalisé, par exemple, en utilisant une interface de réseau qui permet l'introduction d'hypothèses, de règles et de conclusions en langage naturel ou qui les charge à partir de support d'enregistrement (par exemple des disquettes ou des CD-ROM). Un interpréteur assure la traduction d'entrée en langage naturel en règles qui peuvent être exécutées par un ordinateur (par exemple en utilisant un langage tel que Prolog).Advantageously, the system 10 further comprises an interface 24 enrichment in medical rules of memory 18 for storing rules medical. Medical rules can be stored on memory 18 before to use system 10 on several patients; memory 18 can also be updated with new medical rules at each stage of the study of a patient. The new rules are for example loaded from a network or at from archiving bodies such as a CD-ROM. This can be achieved, by example, using a network interface that allows introduction of assumptions, rules and conclusions in natural language or which instructs them to from recording media (eg floppy disks or CD-ROMs). An interpreter provides input translation into natural language into rules which can be executed by a computer (for example using a language such than Prolog).

Le système peut comprendre en outre un module 26 de déduction de règles médicales à partir de la comparaison entre la vérité sur la pathologie obtenue a posteriori - après analyse anatomo-pathologique par exemple - et les données en entrée de l'interface 12 et en sortie de l'unité 20, le module 26 pouvant fournire les règles médicales à la mémoire 18 de stockage de règles médicales. On peut utiliser par exemple, un mécanisme d'apprentissage supervisé qui, à partir de données sur le patient et de la vérité obtenue a posteriori, permet de mettre à jour les règles de décision de la structure de décision arborescente. En particulier, on peut utiliser un réseau de neurones artificiel de type perceptron pour lequel l'apprentissage permet de mettre à jour le poids de ses connexions. Les nouvelles règles construites complètent la mémoire 18 de stockage de règles médicales.The system may further include a rule deduction module 26 medical from the comparison between the truth about the pathology obtained has posteriori - after anatomo-pathological analysis for example - and the data in input of interface 12 and output of unit 20, module 26 being able to supply medical rules in memory 18 for storing medical rules. We can use for example a supervised learning mechanism which, from patient data and the truth obtained a posteriori, allows to update the decision rules of the tree decision structure. In particular, we can use an artificial perceptron neural network for which learning makes it possible to update the weight of its connections. The news constructed rules supplement the memory 18 for storing medical rules.

Le système trouve, par exemple, une application en mammographie.The system finds, for example, an application in mammography.

Claims (8)

Un système numérique d'assistance médicale à la prise de décision, comprenant : une interface (12) de réception recevant : des données cliniques (14), et des données d'enregistrement (16) physiques ou physiologiques, une mémoire (18) de stockage de règles médicales, une unité (20) d'aide à la détection recevant les données d'enregistrement (16) et qui fournit des données (17) mettant en évidence une pathologie, un moteur d'inférences (22) recevant : les données cliniques (14) de l'interface (12), les règles médicales de la mémoire (18) de stockage des règles médicales, et les données (17) fournies par l'unité (20) d'aide à la détection, le moteur (22) fournissant une recommandation à la prise de décision. A digital system of medical assistance for decision-making, including: a reception interface (12) receiving: clinical data (14), and physical or physiological recording data (16), a memory (18) for storing medical rules, a detection aid unit (20) receiving the recording data (16) and which provides data (17) highlighting a pathology, an inference engine (22) receiving: clinical data (14) of the interface (12), the medical rules of the memory (18) for storing the medical rules, and the data (17) provided by the detection assistance unit (20), the engine (22) providing a recommendation for decision-making. Le système de la revendication 1, caractérisé en ce que les données cliniques (14) comprennent le sexe, la taille, le poids, l'âge, l'histoire familiale et les antécédents cliniques d'un patient.The system of claim 1, characterized in that the clinical data (14) includes the sex, height, weight, age, family history and clinical history of a patient. Le système de la revendication 1 ou 2, caractérisé en ce que les données d'enregistrement (16) comprennent des données issues d'un appareil d'imagerie médicale.The system of claim 1 or 2, characterized in that the recording data (16) includes data from a medical imaging device. Le système de la revendication 3, caractérisé en ce que les données (17) comprennent un index de malignité.The system of claim 3, characterized in that the data (17) includes a malignancy index. Le système de la revendication 3 ou 4, caractérisé en ce que les données (17) comprennent une localisation de lésions sur l'organe d'un patient. The system of claim 3 or 4, characterized in that the data (17) includes a location of lesions on the organ of a patient. Le système de l'une des revendications 3 à 5, caractérisé en ce que les données (17) comprennent les caractéristiques de lésions sur l'organe d'un patient.The system of one of claims 3 to 5, characterized in that the data (17) includes the characteristics of lesions on the organ of a patient. Le système de l'une des revendications 1 à 6, caractérisé en ce que le système comprend en outre une interface (24) d'enrichissement en règles médicales de la mémoire (18) de stockage de règles médicales.The system of one of claims 1 to 6, characterized in that the system further comprises an interface (24) for enriching in medical rules the memory (18) for storing medical rules. Le système de l'une des revendications 1 à 7, caractérisé en ce que le système comprend en outre un module (26) de déduction de règles médicales à partir de la comparaison entre la vérité sur la pathologie obtenue a posteriori et les données en entrée de l'interface (12) et en sortie de l'unité (20), le module (26) fournissant les règles médicales à la mémoire (18) de stockage de règles médicales.The system of one of claims 1 to 7, characterized in that the system further comprises a module (26) for deducing medical rules from the comparison between the truth about the pathology obtained a posteriori and the input data from the interface (12) and at the output of the unit (20), the module (26) supplying the medical rules to the memory (18) for storing medical rules.
EP02077665A 2002-07-04 2002-07-04 Digital medical assistance system Withdrawn EP1378853A1 (en)

Priority Applications (3)

Application Number Priority Date Filing Date Title
EP02077665A EP1378853A1 (en) 2002-07-04 2002-07-04 Digital medical assistance system
JP2003270543A JP4651271B2 (en) 2002-07-04 2003-07-03 Computer-aided patient diagnosis decision support system
JP2003192106A JP2004049909A (en) 2002-07-04 2003-07-04 Computer-aided patient health care decision support system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
EP02077665A EP1378853A1 (en) 2002-07-04 2002-07-04 Digital medical assistance system

Publications (1)

Publication Number Publication Date
EP1378853A1 true EP1378853A1 (en) 2004-01-07

Family

ID=29719754

Family Applications (1)

Application Number Title Priority Date Filing Date
EP02077665A Withdrawn EP1378853A1 (en) 2002-07-04 2002-07-04 Digital medical assistance system

Country Status (2)

Country Link
EP (1) EP1378853A1 (en)
JP (2) JP4651271B2 (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB2435950A (en) * 2006-03-10 2007-09-12 Gen Electric Exercise test interpretation
EP3270331A1 (en) 2016-07-15 2018-01-17 Little Brain NV A medical consultation support tool

Families Citing this family (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP4547173B2 (en) * 2004-03-17 2010-09-22 シスメックス株式会社 Diabetes care support system
US20090226065A1 (en) * 2004-10-09 2009-09-10 Dongqing Chen Sampling medical images for virtual histology
JP5133505B2 (en) * 2005-06-24 2013-01-30 ジーイー・メディカル・システムズ・グローバル・テクノロジー・カンパニー・エルエルシー Image determination apparatus and X-ray CT apparatus
EP1946235B1 (en) * 2005-10-31 2018-09-12 Koninklijke Philips N.V. System and method for clinical workflow management
CN101903887B (en) * 2007-12-21 2014-06-11 皇家飞利浦电子股份有限公司 Detection of errors in the inference engine of a clinical decision support system
KR101295613B1 (en) * 2010-03-30 2013-08-12 서울대학교산학협력단 Implementing method for clinical decision support systems of heathcare knowledge with numerous clinical process
CN105144179B (en) 2013-01-29 2019-05-17 分子健康股份有限公司 System and method for clinical decision support
DE102014002173A1 (en) * 2014-02-19 2015-08-20 Dräger Medical GmbH Method for signal and data evaluation for the determination of warning signals for pathological conditions as well as corresponding devices and systems
KR101945459B1 (en) * 2017-07-07 2019-02-07 (의)삼성의료재단 Apparatus, method, and computer program for measuring eye dryness risk

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5769074A (en) * 1994-10-13 1998-06-23 Horus Therapeutics, Inc. Computer assisted methods for diagnosing diseases
WO2000041613A2 (en) 1999-01-14 2000-07-20 Point Loma Industries, Inc. Expert system for real-time decision support
EP1035507A1 (en) 1999-03-12 2000-09-13 Ge Medical Systems Sa Method for detection of elements of interest in a digital radiographic image
WO2001026026A2 (en) * 1999-10-04 2001-04-12 University Of Florida Local diagnostic and remote learning neural networks for medical diagnosis
WO2001069513A2 (en) 2000-03-10 2001-09-20 David S Zakim A system and method for obtaining, processing and evaluating patient information for diagnosing disease and selecting treatment

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP3085724B2 (en) * 1991-05-10 2000-09-11 株式会社東芝 Medical diagnosis support system
JP3365579B2 (en) * 1993-06-10 2003-01-14 株式会社ハーディ Medical diagnostic system
JP3344540B2 (en) * 1996-05-24 2002-11-11 株式会社東芝 Expert system
EP1068568A4 (en) * 1998-04-03 2004-10-27 Triangle Pharmaceuticals Inc Systems, methods and computer program products for guiding the selection of therapeutic treatment regimens
JP4173624B2 (en) * 2000-08-31 2008-10-29 富士フイルム株式会社 Method and apparatus for determining benign / malignant for abnormal shadow candidates

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5769074A (en) * 1994-10-13 1998-06-23 Horus Therapeutics, Inc. Computer assisted methods for diagnosing diseases
WO2000041613A2 (en) 1999-01-14 2000-07-20 Point Loma Industries, Inc. Expert system for real-time decision support
EP1035507A1 (en) 1999-03-12 2000-09-13 Ge Medical Systems Sa Method for detection of elements of interest in a digital radiographic image
WO2001026026A2 (en) * 1999-10-04 2001-04-12 University Of Florida Local diagnostic and remote learning neural networks for medical diagnosis
WO2001069513A2 (en) 2000-03-10 2001-09-20 David S Zakim A system and method for obtaining, processing and evaluating patient information for diagnosing disease and selecting treatment

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB2435950A (en) * 2006-03-10 2007-09-12 Gen Electric Exercise test interpretation
US7542955B2 (en) 2006-03-10 2009-06-02 The General Electric Company Exercise test interpretation
EP3270331A1 (en) 2016-07-15 2018-01-17 Little Brain NV A medical consultation support tool

Also Published As

Publication number Publication date
JP2004130090A (en) 2004-04-30
JP2004049909A (en) 2004-02-19
JP4651271B2 (en) 2011-03-16

Similar Documents

Publication Publication Date Title
Schwendicke et al. Convolutional neural networks for dental image diagnostics: A scoping review
WO2018165620A1 (en) Systems and methods for clinical image classification
KR102296099B1 (en) Tuned medical ultrasound imaging
US8652039B2 (en) Guiding differential diagnosis through information maximization
JP2022502150A (en) Devices and methods for diagnosing gastric lesions using deep learning of gastroscopy images
Chauhan et al. An overview of image processing for dental diagnosis
CN108475540A (en) For using the information from process to carry out the system and method description of dissection modeling
JP7080932B2 (en) Methods and systems for workflow management
US20210398676A1 (en) Machine learning algorithms for detecting medical conditions, related systems, and related methods
Lucas et al. Wound size imaging: ready for smart assessment and monitoring
Scope et al. Correlation of dermoscopic structures of melanocytic lesions to reflectance confocal microscopy
US11232570B2 (en) System and method for diagnosing severity of gastritis
EP1378853A1 (en) Digital medical assistance system
Al Turkestani et al. Clinical decision support systems in orthodontics: a narrative review of data science approaches
Madhumathy et al. Deep learning based photo acoustic imaging for non-invasive imaging
Jaiswal et al. Multi oral disease classification from panoramic radiograph using transfer learning and XGBoost
JP2022126373A (en) Information processing device, information processing method, computer program, and medical diagnostic system
Kim et al. Tooth-Related Disease Detection System Based on Panoramic Images and Optimization Through Automation: Development Study
Jaiswal et al. An intelligent deep network for dental medical image processing system
EP3320462A1 (en) Medical diagnosis assistance method
WO2022201729A1 (en) Image diagnosing system and image diagnosing method
WO2005059802A2 (en) System and method for processing information relating to medical hospital consultations
Silva et al. ProstaTest: A System for the Diagnosis of Prostate Cancer Based on IPSS, PSA, and Ultrasound Scan
US20240005479A1 (en) System and method for cardiovascular risk prediction and computer readable medium thereof
Chadha et al. Signal Processing in Biomedical Applications in Present and Future Development

Legal Events

Date Code Title Description
PUAI Public reference made under article 153(3) epc to a published international application that has entered the european phase

Free format text: ORIGINAL CODE: 0009012

AK Designated contracting states

Kind code of ref document: A1

Designated state(s): AT BE BG CH CY CZ DE DK EE ES FI FR GB GR IE IT LI LU MC NL PT SE SK TR

AX Request for extension of the european patent

Extension state: AL LT LV MK RO SI

17P Request for examination filed

Effective date: 20040707

AKX Designation fees paid

Designated state(s): AT BE BG CH CY CZ DE DK EE ES FI FR GB GR IE IT LI LU MC NL PT SE SK TR

17Q First examination report despatched

Effective date: 20041220

18D Application deemed to be withdrawn

Effective date: 20150202

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: THE APPLICATION IS DEEMED TO BE WITHDRAWN

R18D Application deemed to be withdrawn (corrected)

Effective date: 20150203